Improving policymaking with Big Data
Source: By Jayachandran: Mint
Tucked away in the statement, put out by NITI Aayog after the third meeting of its governing council was an innocuous phrase. The Aayog, the statement said, has partnered with top-ranking institutions “to nurture evidence-based policymaking”. That seems like it should be a given. Surely, all policy is or should be based on solid data. But it became a guiding philosophy of policymaking globally only in the 1990s. It has evolved since then as technological advances have allowed more data to be captured and analysed. To say evidence-based policymaking (EBPM) today is to say policymaking guided by Big Data.
Governance in India, starting with the United Progressive Alliance and continuing into the Narendra Modi administration, has been trending in this direction. The Aadhaar programme, with its hundreds of millions of data points that can be mined for policy formulation and implementation, is a prime example. So is the Centre’s push, kick-started by the currency-swap initiative, to reduce the size of the shadow economy and widen the tax base. The actual amount of black money netted by the initiative might still be a matter of conjecture and debate, but the entire exercise has generated data that can make it tougher for individuals to evade taxes in the future. Geo-tagging of Mahatma Gandhi National Rural Employment Guarantee Scheme assets is another case in point.
This is a natural evolution of EBPM. The spread of the Internet and the rise of social media and the Internet of Things mean that the volume of data we generate is growing exponentially—currently about 2.5 quintillion bytes daily. This is a goldmine for the private sector and governments alike. Online searches can be trawled for data that helps predict disease outbreaks. Cellphone data can help direct relief efforts in the aftermath of a natural disaster. Power-usage data can be analysed to optimize energy grids and plant power generation; discoms in India are already using data from last-mile sensors to implement measures for cutting down aggregated technical and commercial losses.
The uses can range from the national—using healthcare data to revamp the public health system—to the local, where the massive amounts of data generated by cities, from traffic signals to public transport usage, can be used to improve infrastructure and transport systems as Singapore has done. But utilizing Big Data effectively in this fashion will mean keeping a few factors in mind.
First, two of the defining characteristics of Big Data analytics are the volume and veracity of data. The computing concept of garbage in, garbage out holds true here. Infrastructure in India for efficient data collection and management is lacking; this must be strengthened. The comptroller and auditor general’s Big Data management policy and its establishment of the Centre for Data Management and Analytics is positive signs in this context; they show that such issues are on the radar.
Volume and veracity also necessitate sharing of data across ministries and departments—indeed, with the public at large—to allow private-sector solutions that can in turn be utilized in government policymaking. India has done well to join the Open Government Data movement and formulated a national data sharing and accessibility policy in 2012. The government’s Open Government Data portal is a significant step, with thousands of data sets available regarding everything from health to agriculture. However, the reliability of the data, the tendency to work in silos, the reluctance to share data and lack of standard formats are all concerns.
Second, another marker of Big Data analytics is data velocity. Large amounts of data are collected swiftly today; this also means that much of it loses relevance after some time. Using Big Data effectively for policy formulation will thus mean changing policymaking structures and processes—continuously re-evaluating and rejigging policies based on the feedback generated by new data, from on the ground results to public opinion scraped from social media. Incorporating this flexibility in hierarchical bureaucratic structures will not be easy.
Third, the ethics of Big Data analytics is an area of major debate. The issues range from anonymization of data to what data should be collected and what use it should be put to. These issues will loom larger as new fields like psychometrics—the combination of Big Data with behavioural science to determine various aspects of people’s lives—evolves. The Indian state must engage robustly with these issues. This is currently lacking given the failure to enact even basic laws about data privacy and the right to privacy.
Fourth, and perhaps most importantly, the right dynamic between political imperatives and EBPM will have to be struck. When the latter gained currency in the UK in the 1990s the government of the day touted it as a “post-ideological” approach to policymaking. This is a chimera. All the data in the world will not replace the political process; the farm-loan waiver mess is a case in point. While Big Data can and should be used to inform policymaking, the biases and motives of the political process that guide its usage should not be forgotten.
When the Planning Commission was created in 1950, early planners such as P.C. Mahalanobis realized the importance of good data. Consequently, India had one of the most effective data systems going at the time. The quality of data collection and usage has declined since. NITI Aayog’s push to EBPM is thus timely. But it is only the beginning of a long process.
